A case based reasoning model for multilingual language generation in dialogues

The process of Natural Language Generation for a Conversational Agent translates some semantic language to its surface form expressed in natural language. In this paper, we are going to show a Case Based Reasoning technique which is easily extensible and adaptable to multiple domains and languages, that generates coherent phrases and produces a natural outcome in the context of a Conversational Agent that maintains a dialogue with the user.     

Comments on “A novel dead time compensator for stable processes with long dead times” [Journal of Process Control 22 (2012) 612–625]

This short communication analyzes the results recently presented in the paper “On a novel dead time compensator for stable processes with long dead times” published in the Journal of Process Control. In the mentioned paper it is argued that the proposed strategy, called modified Smith predictor (MSP), gives better performance than the filtered Smith predictor (FSP) dead-time compensator for stable processes with dead time. In fact MSP has the same structure as FSP and only some specific tuning rules of the filters are proposed. Therefore, in this work some aspects of the comparative analysis and tuning rules presented in the referred paper are discussed to show that MSP is a particular case of FSP and that for some particular cases its tuning rule does not allow for a good closed-loop response. Some simulation case studies are used to illustrate these aspects.     

Chance-constrained model predictive control for spacecraft rendezvous with disturbance estimation

A robust Model Predictive Controller (MPC) is used to solve the problem of spacecraft rendezvous, using the Hill-Clohessy-Wiltshire model with additive disturbances and line-of-sight constraints. Since a standard (non-robust) MPC is not able to cope with disturbances, a robust MPC is designed using a chance-constrained approach for robust satisfaction of constraints in a probabilistic sense. Disturbances are modeled as Gaussian allowing for an explicit transformation of the probabilistic constraints into simple algebraic constraints. To estimate the distribution parameters a predictor of disturbances is proposed. Both robust and non-robust MPC control laws are compared using the Monte Carlo method, which shows the superiority of the robust MPC.     

A learning result for continuous-time recurrent neural networks

The following learning problem is considered, for continuous-time recurrent neural networks having sigmoidal activation functions. Given a “black box” representing an unknown system, measurements of output derivatives are collected, for a set of randomly generated inputs, and a network is used to approximate the observed behavior. It is shown that the number of inputs needed for reliable generalization (the sample complexity of the learning problem) is upper bounded by an expression that grows polynomially with the dimension of the network and logarithmically with the number of output derivatives being matched.     

Packet Video Broadcasting with General-Purpose Operating Systems in an Ethernet

Video transmission with general-purpose PCs poses a number of requirements that radically differ from those of high-end dedicated video servers. We analyze the scenario of an Ethernet local area network in which a number of PCs are transmitting video streams, while other TCP/IP applications are also running concurrently. Our findings show that since the operating system clock resolution cannot cope with the transmission timing requirements the following holds: if the video transmission is performed with exact timing accuracy to maintain a constant rate then CPU load grows to 100%, thus blocking the PC for other user applications; on the other hand, if transmission is performed in a bursty manner, i.e. with sleep system calls, then CPU load decreases dramatically but the increased burstiness of the video stream has a negative impact on network performance (for example, capture effect in the Ethernet). Furthermore, the impact of video transmission over the rest of TCP/IP applications running on the same network depends heavily on the packet size. We provide an integrated analysis of operating system and network parameters to achieve video broadcasting while preserving timing requirements and minimizing the impact on other applications.     

Fine structure of the male accessory glands of Triatoma rubrofasciata (De Geer, 1773) (Hemiptera, Triatominae)

Male of Triatoma rubrofasciata has four elongated sac-like reproductive mesodermic accessory glands, lined by an inner single layer of secretory cells, with basal plasma membrane infolds and short apical microvilli, and externally enveloped by a thin visceral muscle layer. The secretory cells have a well-developed rough endoplasmic reticulum, Golgi complex, mitochondria, and secretory granules. In one day old adult the gland cells are poorly developed, presenting small, electron-transparent secretory granules scattered among the rough endoplasmatic reticulum, whereas in three days old adult these cells have the cisternae of the rough endoplasmatic reticulum varing size degree, filled with granular electrondense content. In five days old males the secretory granules increase in diameter, being released to the gland lumen. Therefore, there is an increase of the secretory activity according to male maturation.     

Learning Structure from Data and Its Application to Ozone Prediction

In this paper we propose an algorithm for structure learning in predictive expert systems based on a probabilistic network representation. The idea is to have the simplest structure (minimum number of links) with acceptable predictive capability. The algorithm starts by building a tree structure based on measuring mutual information between pairs of variables, and then it adds links as necessary to obtain certain predictive performance. We have applied this method for ozone prediction in México City, where the ozone level is used as a global indicator for the air quality in different parts of the city. It is important to predict the ozone level a day, or at least several hours in advance, to reduce the health hazards and industrial losses that occur when the ozone reaches emergency levels. We obtained as a first approximation a tree-structured dependency model for predicting ozone in one part of the city. We observe that even with only three parameters, its estimations are acceptable.A causal network representation and the structure learning techniques produced some very interesting results for the ozone prediction problem. Firstly, we got some insight into the dependence structure of the phenomena. Secondly, we got an indication of which are the important and not so important variables for ozone forecasting. Taking this into account, the measurement and computational costs for ozone prediction could be reduced. And thirdly, we have obtained satisfactory short term ozone predictions based on a small set of the most important parameters.     

On characterizations of the input-to-state stability property

We show that the well-known Lyapunov sufficient condition for “input-to-state stability” (ISS) is also necessary, settling positively an open question raised by several authors during the past few years. Additional characterizations of the ISS property, including one in terms of nonlinear stability margins, are also provided.     

The Stress Response of the Holothurian Central Nervous System: A Transcriptomic Analysis

Injury to the central nervous system (CNS) results in permanent damage and lack of function in most vertebrate animals, due to their limited regenerative capacities. In contrast, echinoderms can fully regenerate their radial nerve cord (RNC) following transection, with little to no scarring. Investigators have associated the regenerative capacity of some organisms to the stress response and inflammation produced by the injury. Here, we explore the gene activation profile of the stressed holothurian CNS. To do this, we performed RNA sequencing on isolated RNC explants submitted to the stress of transection and enzyme dissection and compared them with explants kept in culture for 3 days following dissection. We describe stress-associated genes, including members of heat-shock families, ubiquitin-related pathways, transposons, and apoptosis that were differentially expressed. Surprisingly, the stress response does not induce apoptosis in this system. Other genes associated with stress in other animal models, such as hero proteins and those associated with the integrated stress response, were not found to be differentially expressed either. Our results provide a new viewpoint on the stress response in the nervous system of an organism with amazing regenerative capacities. This is the first step in deciphering the molecular processes that allow echinoderms to undergo fully functional CNS regeneration, and also provides a comparative view of the stress response in other organisms.     

Study of the Ghrelin/LEAP-2 Ratio in Humans and Rats during Different Phases of Pregnancy

The Liver-Expressed Antimicrobial Peptide 2 (LEAP-2) has emerged as an endogenous GHS-R antagonist and blunts the orexigenic action of ghrelin. This study aimed to determine the Ghrelin/LEAP-2 ratio in humans and rats during pregnancy. In humans, we conducted a nested case-control study within an observational prospective cohort. Healthy and mild preeclamptic pregnant women were studied at each trimester of gestation and three months postpartum. In addition, a group of non-pregnant women was studied into the follicular and luteal phases of the menstrual cycle. Furthermore, Ghrelin/LEAP-2 ratio was investigated in non-pregnant rats and at different periods of rat pregnancy. Human and rat serum ghrelin and LEAP-2 levels were determined using the commercially available ELISA kits. The Ghrelin/LEAP-2 ratio peak around the second trimester of gestation in healthy pregnant women (p < 0.05). Additionally, there were no statistically significant differences in Ghrelin/LEAP-2 ratio between healthy and preeclamptic pregnant women at each trimester of gestation (p > 0.05). The Ghrelin/LEAP-2 ratio in pregnant rat reached the peak around mid-gestation with a similar pattern to the human pregnancy. LEAP-2 was visualized by immunohistochemistry in human term placenta and rat placentas on days 12, 16 and 21 of pregnancy. In conclusion, this study provides the first evidence of a Ghrelin/LEAP-2 ratio peak around the half-way point of pregnancy onwards during human and rat pregnancy, and it might be associated with increased rates of weight gain during pregnancy. Thus, this study suggests that LEAP-2 and Ghrelin/LEAP-2 ratio might play an important role in maternal physiology adaptation of weight gain during pregnancy.